agentmods.dev

harness-control AGENTS.md

Project instructions for harness-control, a control system that manages how AI agents work on software projects.

In plain words
What is it for?
Use them when navigating the repository, changing code, checking results, recording architecture decisions, or running evaluations.
Why use it?
They define which context and rules to read, how to verify work, and what claims or commands are forbidden, reducing unsafe or unverified changes.

Instructions file for CodexOpenCode

22 1mo ago C scan 2,192 tokens original MIT
View the source file harness-base/harness-control
Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add instructions/harness-base/harness-control/agents-md
Clone the repo
git clone --depth 1 https://github.com/harness-base/harness-control

Made for: Codex, OpenCode.

Per session 2,192 This file is loaded in full into every session.
When invoked 2,192 The same file — it is already loaded in full.
Security scan C 1 finding. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.02192 $0.02192
Opus 5 $0.01096 $0.01096
Sonnet 5 $0.00438 $0.00438
Haiku 4.5 $0.00219 $0.00219

Measured 3d ago against content hash e745ce6e4468, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade C, and why

harness-control AGENTS.md scanned grade C with 1 finding against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 3d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Recursive force deletehighDestructive command

rm -rf with a variable or a broad path is one typo away from removing the wrong tree.

- **不碰密钥与危险命令**：不泄露密钥 / token；不执行 `git reset --hard`、`rm -rf /` 等高危命令（hook 会拦）。 <!-- rule: rule-0006 | sev: blocker -->
AGENTS.md · 63 lines

How it starts

The opening of the file, as written. The whole thing — 63 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Agent Harness 控制面

本仓是一套通用的 agent 控制面（harness）：用「最小内核 + 可挂载模块」治理 AI agent 在被管工程上的开发。被管工程挂进 projects/（已挂 kratos-base）。

默认策略：少读、按需读、渐进加载。

启动顺序

  1. 读 docs/context/CURRENT_STATUS.md（当前真实状态）。
  2. 按需读、顺引用链走（默认少读；引用链怎么走见 docs/context/CONTEXT_LOADING.md，读什么由你按任务自主判断，不判档）。
  3. 需要文档路由时，读 docs/README.md。
  4. 在某目录读或改代码前，加载该位置向上最近的 AGENTS.md（连同其同级 CLAUDE.md）——就近规则随之生效；按目录加载与渐进式引用叠加，详见 docs/context/CONTEXT_LOADING.md。
  5. 进入某目录想读 / 动其下文件前，若该目录有 README.md，先读一下——不必通读，目的是知道这里有什么、该挑哪个。README 不像 AGENTS.md 那样自动加载，靠这条规则触发。

规则

下面是 harness 全局规则（红线，动手前必看）。编号 rule-00NN 是稳定引用键，被 eval 考题 / ADR 按号引用；全文就在这里。每条带隐形标记供 rules-index 扫描。项目专属规则沉淀在 projects/**/AGENTS.md（就近生效），不堆在这里。（rule-0001 已退役见 ADR-0023、rule-0004 已退役见 ADR-0025——编号永久空缺。）

  • blocked / skipped ≠ pass：验证没真跑通，不许声称通过。
  • 不许假完成：没有真实运行证据，不得声称功能完成或验收通过。
  • 收尾前过 eval：命中任一判据的任务——多步改产物 / 写了 ADR / 动业务代码 / 关键决策点——收尾前必须跑 task eval review（独立评委按 rubric 打分）；轻量问答、琐碎修补不触发。
  • 不碰密钥与危险命令：不泄露密钥 / token；不执行 git reset --hard、rm -rf / 等高危命令（hook 会拦）。
  • 改架构 / 接口须回顾相关 skill：大改（写了 ADR）必须回顾 .agents/skills/，更新或写明无需更新。
  • 外部材料不自动采信：事实源 = 正式文档 + 工程当前代码；外部 / 粘贴材料要先整理验收才算数。
  • 验收断言必须锚定唯一、真实、产出方的证据：断言绑到唯一真实信号（防共因污染 / 防超时竞态掩盖）；声称的保证必须有守护测试；测试不许为通过而牵强、注释不许撒谎。
  • PRD 产出标准：产出 PRD 时——先有 approved 用户故事（独立 user-stories.md 为上游、PRD 与之对齐）、验收可观测、范围 in+out 闭合、每页四态、（若产出）原型可点通、假设显式确认、可追溯、登记不漂移（仅在产出 PRD 时适用，不强制 PRD 必须存在）。
  • 决策与知识必须当轮落文档，落文档提醒兜住遗漏：改了产物或做了关键决策，知识要就近写进 AGENTS.md/lessons/规则/ADR/memory；用户纠正也算（用户说"不是这样 / 你理解错了 / 撤回 / 你搞混了"时，当轮记一条 tasks/lessons.md 三段式：错在哪 / 怎么防 / 怎么更早发现）；遗漏由 Stop hook 的落文档提醒兜底复查、写 tasks/optimization-log.md（机制见 docs/harness/HOOKS.md），捞到的须落到对应文档、不许烂在 log 里。
  • 状态/索引文档不硬编码可自动生成的枚举：凡已有 *-index 自动生成权威清单的（skill→.agents/skills/README.md、规则→docs/rules/index.yaml 等），CURRENT_STATUS 等状态文档只写"以该自动生成索引为准"的指针，不复刻计数/清单——硬编码枚举无 --check 守、每次新增就漂。举 1–2 例可以，整列枚举不行。
  • 非琐碎任务维护 tasks/todo.md（标 eval + 收尾 Review）：多步 / 动产物 / 动业务码的任务，动手即在 tasks/todo.md 立当前任务并标 eval: 要|不要 ｜ task: <名>（按 rule-0005 判据自判；收尾闸据此判要不要 eval）；范围变即更新、收尾前补 Review 段；保持轻——旧块滚进 tasks/archive/，不长成流水账。
  • 测试用例产出标准：产出测试用例时——每条验收点 AC 与每个功能点 FP 都被 ≥1 条用例 covers: 覆盖（无遗漏、无悬空引用，test-cases-audit 硬闸机检）、用例覆盖正常 / 边界 / 异常、covers: 为覆盖关系唯一真相源（不另存手维护映射表）、产物登记不漂移；只管"用例齐不齐 / 覆盖全不全"，不碰"过没过"（执行结果另起）；用例真覆盖语义 / 边界异常齐由 eval 考题 015 判（仅在产出测试用例时适用，不强制必须存在）。
  • 控制面与项目内容隔离（命根）：harness 资产（skill / 模板 / 规则 / 子 agent / 脚本）必须通用、项目无关——不掺任何具体项目的领域名词 / 业务规则 / 多租户（tenant_id）等假设，示例用中性占位（如 GET /v1/items）；具体项目内容只落被管工程 projects/<工程>/ 与产物区（docs/prds·designs·test-cases）。改 harness 资产时守住这条边界：通用的归资产、具体的归产物。
  • 不擅自 git 写操作：未经许可不 commit / push / reset / 删分支 / 改 remote。

Read the full file on GitHub · 63 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 3d ago First seen · 63 lines · 2,192 tokens per session scan C e745ce6e4468

Subscribe to this mod's changes

harness-control AGENTS.md is an instructions file published in the GitHub repository harness-base/harness-control (22 stars, last pushed 1mo ago), licensed MIT. It adds 2,192 tokens to every session, about $0.0110 per session on Opus 5. A static security scan graded it C with 1 finding (recursive force delete). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.

Building agents
Related

Other instructions, from other repositories

vscode buildNext.instructions.md

Working notes and architecture documentation for the new esbuild-based build system in build/next. Use when making changes to the new build pipeline (transpile/bundle commands, NLS plugin, source-map handling, resource copying, or self-hosting watch tasks).

microsoft/vscode · 6,785 tokens

spec-kit AGENTS.md

AGENTS.md instructions for github/spec-kit, covering agents.md, about spec kit and specify, quickstart — add a new integration in 5 steps, integration architecture and integrationmanifest — file tracking.

github/spec-kit · 7,104 tokens

codex AGENTS.md

AGENTS.md instructions for openai/codex, covering rust/codex-rs, the codex-core crate, code review rules, crate api surface and model visible context.

openai/codex · 5,182 tokens

langchain AGENTS.md

AGENTS.md instructions for langchain-ai/langchain, covering global development guidelines for the langchain monorepo, corridor security analysis, project architecture and context, monorepo structure and development tools & commands.

langchain-ai/langchain · 4,345 tokens

vscode oss-third-party-notices.instructions.md

Instructions for microsoft/vscode, covering vs code oss third-party-notices pipeline, architecture, pipeline flow in ci, applying the notice (cutover) and fallback chain (never fail the build).

microsoft/vscode · 5,001 tokens

next.js AGENTS.md

Instructions for vercel/next.js, covering next.js development guide, codebase structure, monorepo overview, core package: packages/next and other important packages.

vercel/next.js · 7,296 tokens